Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!sol.ctr.columbia.edu!news.kei.com!ub!acsu.buffalo.edu!ubvms.cc.buffalo.edu!comjoew
From: comjoew@ubvms.cc.buffalo.edu (Joseph D Woelfel)
Subject: Multidimensional Scaling
Message-ID: <Cz5u6G.159@acsu.buffalo.edu>
News-Software: VAX/VMS VNEWS 1.50AXP 
Sender: nntp@acsu.buffalo.edu
Nntp-Posting-Host: ubvmsa.cc.buffalo.edu
Organization: University at Buffalo
Date: Sat, 12 Nov 1994 15:11:00 GMT
Lines: 20

	Leon Heller suggests using Multidimensional Scaling as a representation
of cognitive structure. CATPAC provides MDS representations of its internal
states. Especially if the net is represented by a square matrix, the
	principle axes of the scalar products matrix fromed from this
	will yield coordinates for plotting each neuron in a MDS space;
the distances between neurons will represent their dissimilarities.

	For a more general approach to representing cognitive structures
by MDS, I recommend "The Measurement of Communication Processes: Galileo
Theory and Method, by Woelfel and Fink, Academic Press, 1980.

	Someone else might suggest you investigate non-metric procedures,
but these have never worked well for me.

Joseph Woelfel
University at Buffalo



 8nkNY t"closenehequal  
